import json
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
import datetime

import seaborn
seaborn.set_theme()

pd.set_option('future.no_silent_downcasting', True)

with open('regions-m49.json') as f:
    regions = json.load(f)['regions']
    regions = {v: n for n, v in regions.items()}

with open('countries-m49.json') as f:
    countries = json.load(f)['countries']
    country_subregions = {c['name']: regions[c['subRegion']] for c in countries}
    country_regions = {c['name']: regions[c['region']] for c in countries}

country_mapping = {
    'Czech Republic': 'Czechia',
    'Iran': 'Iran (Islamic Republic of)',
    'Russia': 'Russian Federation',
    'South Korea': 'Republic of Korea',
    'Taiwan': 'China',  # UN is an abomination
    'Turkey': 'T\u00fcrkiye',
    'United Kingdom': 'United Kingdom of Great Britain and Northern Ireland',
    'United States': 'United States of America',
}

max_year = 2023
years = range(1954, max_year + 1)
max_date = pd.to_datetime(datetime.date(year = max_year + 1, month = 1, day = 1))

copyright_text = 'CC BY-SA 4.0 Andrey Upadyshev (image) and\nWikipedia, List of commercial nuclear reactors (data)'
copyright_font_size = 10

def format_years_ticks(ax):
    for label in ax.get_xticklabels():
        if label.get_text() in ('1957', '1979', '1986', '2011'):
            label.set_color('red')


def format_number_plants_ticks(ax, ymin, ymax):
    def fmt(val, pos):
        return int(abs(val))
    ax.yaxis.set_major_formatter(fmt)
    ax.set_ylim(-ymin, ymax)

orig_input_df = pd.read_csv('./reactors.csv').convert_dtypes()
for col in ('Begin building', 'Commercial operation', 'Closed'):
    orig_input_df[f'{col} date'] = pd.to_datetime(orig_input_df[col])
    orig_input_df[f'{col}'] = orig_input_df[f'{col} date'].dt.year

input_df = orig_input_df[orig_input_df['Begin building'].notna() & (orig_input_df['Begin building'] <= max_year)].reset_index(drop=True)
input_df.loc[input_df['Commercial operation'] > max_year, 'Commercial operation'] = None
input_df.loc[input_df['Closed'] > max_year, 'Closed'] = None

input_df['SubRegion'] = input_df['Country'].map(lambda x: country_subregions[country_mapping.get(x, x)])
input_df['Region'] = input_df['Country'].map(lambda x: country_regions[country_mapping.get(x, x)])
input_df['Operated closed'] = input_df['Closed'].where(input_df['Commercial operation'].notna(), None)
input_df['Construction time'] = (input_df['Commercial operation date'] - input_df['Begin building date']).dt.total_seconds() / (365.25 * 24 * 60 * 60)
input_df

total_constructed_by_subregion = input_df.groupby(['SubRegion', 'Region'])[['Begin building', 'Commercial operation', 'Closed']].count().reset_index().sort_values('Begin building', ascending=False).reset_index(drop=True)
total_constructed_by_subregion

total_constructed_by_country = input_df.groupby(['Country', 'SubRegion', 'Region'])[['Begin building', 'Commercial operation', 'Closed']].count().reset_index().sort_values('Begin building', ascending=False).reset_index(drop=True)
total_constructed_by_country

num_building_started = input_df.groupby(['Begin building'])['Begin building'].count()
num_connected = input_df.groupby(['Commercial operation'])['Commercial operation'].count()
num_closed = input_df.groupby(['Closed'])['Closed'].count()
num_operated_closed = input_df.groupby(['Operated closed'])['Operated closed'].count()

max_building_started = num_building_started.max()
max_num_closed = num_closed.max()
ymax = (max_building_started + 4) // 5 * 5
ymin = (max_num_closed + 4) // 5 * 5

combined = pd.concat([
        num_building_started,
        num_connected,
        num_closed,
        num_operated_closed,
    ], axis=1)

combined = combined.reindex(years, fill_value=0)

combined['Commercial operation tot'] = combined['Commercial operation'].fillna(0).cumsum()
combined['Operated closed tot'] = combined['Operated closed'].fillna(0).cumsum()
combined['In operation'] = combined['Commercial operation tot'] - combined['Operated closed tot']

fig, ax = plt.subplots(2, 1, figsize=(11, 10))

ax1, ax2 = ax

combined['Begin building'].plot.bar(ax=ax1, label='Construction started')
(-combined['Closed']).plot.bar(ax=ax1, color='red', label='Closed')

title = 'World'
print(f'{title}:')
ax1.set_title(title)
ax1.set_xlabel(None)
format_years_ticks(ax1)
format_number_plants_ticks(ax1, ymin, ymax)
ax1.text(0, -ymin, copyright_text, fontsize=copyright_font_size, verticalalignment='bottom')
ax1.legend()

combined['In operation'].plot.bar(ax=ax2, label='Number in operation')
combined['Commercial operation'].plot.bar(ax=ax2, color='black', label='Operation started')
format_years_ticks(ax2)
#ax2.text(0, ax2.get_ylim()[1], copyright_text, fontsize=copyright_font_size, verticalalignment='top')
ax2.legend()

fig.tight_layout()
plt.show()

World:

total_constructed_by_region = input_df.groupby(['Region'])[['Begin building', 'Commercial operation', 'Closed']].count().reset_index().sort_values('Begin building', ascending=False).reset_index(drop=True)

num_building_started = input_df.groupby(['Region', 'Begin building'])['Begin building'].count()
num_connected = input_df.groupby(['Region', 'Commercial operation'])['Commercial operation'].count()
num_closed = input_df.groupby(['Region', 'Closed'])['Closed'].count()
num_operated_closed = input_df.groupby(['Region', 'Operated closed'])['Operated closed'].count()

#top_regions = total_constructed_by_region[total_constructed_by_region['Begin building'] > 10]['Region'].to_list()
top_regions = total_constructed_by_region['Region'].to_list()

regions = sorted(c for c in input_df['Region'].unique() if c in num_building_started)

max_building_started = num_building_started.max()
max_num_closed = num_closed.max()
ymax = (max_building_started + 4) // 5 * 5
ymin = (max_num_closed + 4) // 5 * 5
tot_ymax_regions = 250  # TODO: calculate from data


for c in top_regions:
    combined = pd.concat([
        num_building_started[c],
        num_connected[c] if c in num_connected else pd.DataFrame(columns=['Commercial operation']),
        num_closed[c] if c in num_closed else pd.DataFrame(columns=['Closed']),
        num_operated_closed[c] if c in num_operated_closed else pd.DataFrame(columns=['Operated closed']),
    ], axis=1)
    
    combined = combined.reindex(years, fill_value=0)

    combined['Commercial operation tot'] = combined['Commercial operation'].fillna(0).cumsum()
    combined['Operated closed tot'] = combined['Operated closed'].fillna(0).cumsum()
    combined['In operation'] = combined['Commercial operation tot'] - combined['Operated closed tot']

    fig, ax = plt.subplots(2, 1, figsize=(11, 10))

    ax1, ax2 = ax

    combined['Begin building'].plot.bar(ax=ax1, label='Construction started')
    (-combined['Closed']).plot.bar(ax=ax1, color='red', label='Closed')

    title = f'{c} (UN M49)'
    print(f'{title}:')
    ax1.set_title(title)
    ax1.set_xlabel(None)
    format_years_ticks(ax1)
    format_number_plants_ticks(ax1, ymin, ymax)
    ax1.text(0, -ymin, copyright_text, fontsize=copyright_font_size, verticalalignment='bottom')
    ax1.legend()

    combined['In operation'].plot.bar(ax=ax2, label='Number in operation')
    combined['Commercial operation'].plot.bar(ax=ax2, color='black', label='Operation started')
    format_years_ticks(ax2)
    ax2.set_ylim(0, tot_ymax_regions)
    #ax2.text(0, ax2.get_ylim()[1], copyright_text, fontsize=copyright_font_size, verticalalignment='top')
    ax2.legend()


    fig.tight_layout()
    plt.show()

Europe (UN M49):

Asia (UN M49):

Americas (UN M49):

Africa (UN M49):

europe_df = input_df[(input_df['Region'] == 'Europe') & (input_df['SubRegion'] != 'Eastern Europe') & (input_df['SubRegion'] != 'Lithuania')]
num_building_started = europe_df.groupby(['Begin building'])['Begin building'].count()
num_connected = europe_df.groupby(['Commercial operation'])['Commercial operation'].count()
num_closed = europe_df.groupby(['Closed'])['Closed'].count()
num_operated_closed = europe_df.groupby(['Operated closed'])['Operated closed'].count()

max_building_started = num_building_started.max()
max_num_closed = num_closed.max()
ymax = (max_building_started + 4) // 5 * 5
ymin = (max_num_closed + 4) // 5 * 5

combined = pd.concat([
        num_building_started,
        num_connected,
        num_closed,
        num_operated_closed,
    ], axis=1)

combined = combined.reindex(years, fill_value=0)

combined['Commercial operation tot'] = combined['Commercial operation'].fillna(0).cumsum()
combined['Operated closed tot'] = combined['Operated closed'].fillna(0).cumsum()
combined['In operation'] = combined['Commercial operation tot'] - combined['Operated closed tot']

fig, ax = plt.subplots(2, 1, figsize=(11, 10))

ax1, ax2 = ax

combined['Begin building'].plot.bar(ax=ax1, label='Construction started')
(-combined['Closed']).plot.bar(ax=ax1, color='red', label='Closed')

title = 'Europe excluding Eastern Europe and ex-USSR countries'
print(f'{title}:')
ax1.set_title(title)
ax1.set_xlabel(None)
format_years_ticks(ax1)
format_number_plants_ticks(ax1, ymin, ymax)
ax1.text(0, -ymin, copyright_text, fontsize=copyright_font_size, verticalalignment='bottom')
ax1.legend()

combined['In operation'].plot.bar(ax=ax2, label='Number in operation')
combined['Commercial operation'].plot.bar(ax=ax2, color='black', label='Operation started')
format_years_ticks(ax2)
ax2.set_ylim(0, tot_ymax_regions)
#ax2.text(0, ax2.get_ylim()[1], copyright_text, fontsize=copyright_font_size, verticalalignment='top')
ax2.legend()

fig.tight_layout()
plt.show()

Europe excluding Eastern Europe and ex-USSR countries:

num_building_started = input_df.groupby(['SubRegion', 'Begin building'])['Begin building'].count()
num_connected = input_df.groupby(['SubRegion', 'Commercial operation'])['Commercial operation'].count()
num_closed = input_df.groupby(['SubRegion', 'Closed'])['Closed'].count()
num_operated_closed = input_df.groupby(['SubRegion', 'Operated closed'])['Operated closed'].count()

sub_regions = sorted(c for c in input_df['SubRegion'].unique() if c in num_building_started)
#top_regions = total_constructed_by_subregion[total_constructed_by_subregion['Begin building'] > 10]['SubRegion'].to_list()
top_regions = total_constructed_by_subregion['SubRegion'].to_list()

max_building_started = num_building_started.max()
max_num_closed = num_closed.max()
ymax = (max_building_started + 4) // 5 * 5
ymin = (max_num_closed + 4) // 5 * 5


for c in top_regions:
    combined = pd.concat([
        num_building_started[c],
        num_connected[c] if c in num_connected else pd.DataFrame(columns=['Commercial operation']),
        num_closed[c] if c in num_closed else pd.DataFrame(columns=['Closed']),
        num_operated_closed[c] if c in num_operated_closed else pd.DataFrame(columns=['Operated closed']),
    ], axis=1)
    
    combined = combined.reindex(years, fill_value=0)

    combined['Commercial operation tot'] = combined['Commercial operation'].fillna(0).cumsum()
    combined['Operated closed tot'] = combined['Operated closed'].fillna(0).cumsum()
    combined['In operation'] = combined['Commercial operation tot'] - combined['Operated closed tot']

    fig, ax = plt.subplots(2, 1, figsize=(11, 10))

    ax1, ax2 = ax

    combined['Begin building'].plot.bar(ax=ax1, label='Construction started')
    (-combined['Closed']).plot.bar(ax=ax1, color='red', label='Closed')

    title = f'{c} (UN M49)'
    print(f'{title}:')
    ax1.set_title(title)
    ax1.set_xlabel(None)
    format_years_ticks(ax1)
    format_number_plants_ticks(ax1, ymin, ymax)
    ax1.text(0, -ymin, copyright_text, fontsize=copyright_font_size, verticalalignment='bottom')
    ax1.legend()

    combined['In operation'].plot.bar(ax=ax2, label='Number in operation')
    combined['Commercial operation'].plot.bar(ax=ax2, color='black', label='Operation started')
    format_years_ticks(ax2)
    ax2.set_ylim(0, tot_ymax_regions)
    #ax2.text(0, ax2.get_ylim()[1], copyright_text, fontsize=copyright_font_size, verticalalignment='top')
    ax2.legend()


    fig.tight_layout()
    plt.show()

Eastern Asia (UN M49):

Northern America (UN M49):

Western Europe (UN M49):

Eastern Europe (UN M49):

Northern Europe (UN M49):

Southern Asia (UN M49):

Southern Europe (UN M49):

Latin America and the Caribbean (UN M49):

Western Asia (UN M49):

Northern Africa (UN M49):

num_building_started = input_df.groupby(['Country', 'Begin building'])['Begin building'].count()
num_connected = input_df.groupby(['Country', 'Commercial operation'])['Commercial operation'].count()
num_closed = input_df.groupby(['Country', 'Closed'])['Closed'].count()
num_operated_closed = input_df.groupby(['Country', 'Operated closed'])['Operated closed'].count()

countries = sorted(c for c in input_df['Country'].unique() if c in num_building_started)
#top_countries = total_constructed_by_country[total_constructed_by_country['Begin building'] >= 10]['Country'].to_list()
top_countries = total_constructed_by_country['Country'].to_list()

max_building_started = num_building_started.max()
max_num_closed = num_closed.max()
ymax = (max_building_started + 4) // 5 * 5
ymin = (max_num_closed + 4) // 5 * 5
tot_ymax = 120  # TODO: calculate from data


for c in top_countries:
    combined = pd.concat([
        num_building_started[c],
        num_connected[c] if c in num_connected else pd.DataFrame(columns=['Commercial operation']),
        num_closed[c] if c in num_closed else pd.DataFrame(columns=['Closed']),
        num_operated_closed[c] if c in num_operated_closed else pd.DataFrame(columns=['Operated closed']),
    ], axis=1)
    
    combined = combined.reindex(years, fill_value=0)

    combined['Commercial operation tot'] = combined['Commercial operation'].fillna(0).cumsum()
    combined['Operated closed tot'] = combined['Operated closed'].fillna(0).cumsum()
    combined['In operation'] = combined['Commercial operation tot'] - combined['Operated closed tot']

    fig, ax = plt.subplots(2, 1, figsize=(11, 10))

    ax1, ax2 = ax

    combined['Begin building'].plot.bar(ax=ax1, label='Construction started')
    (-combined['Closed']).plot.bar(ax=ax1, color='red', label='Closed')

    title = c
    print(f'{title}:')
    ax1.set_title(title)
    ax1.set_xlabel(None)
    format_years_ticks(ax1)
    format_number_plants_ticks(ax1, ymin, ymax)
    ax1.text(0, -ymin, copyright_text, fontsize=copyright_font_size, verticalalignment='bottom')
    ax1.legend()

    combined['In operation'].plot.bar(ax=ax2, label='Number in operation')
    combined['Commercial operation'].plot.bar(ax=ax2, color='black', label='Operation started')
    format_years_ticks(ax2)
    ax2.set_ylim(0, tot_ymax)
    #ax2.text(0, ax2.get_ylim()[1], copyright_text, fontsize=copyright_font_size, verticalalignment='top')
    ax2.legend()

    fig.tight_layout()
    plt.show()

United States:

China:

France:

Japan:

Russia:

United Kingdom:

Germany:

India:

South Korea:

Canada:

# Never opened
never_opened_df = input_df[input_df['Commercial operation'].isna() & input_df['Begin building'].notna() & input_df['Closed'].notna()].sort_values('Closed')

num_building_started = never_opened_df.groupby(['Begin building'])['Begin building'].count()
num_closed = never_opened_df.groupby(['Closed'])['Closed'].count()

combined = pd.concat([
        num_building_started,
        num_closed,
    ], axis=1)

combined = combined.reindex(years, fill_value=0)

fig, ax = plt.subplots(figsize=(11, 5))

combined['Begin building'].plot.bar(ax=ax, label='Construction started')
(-combined['Closed']).plot.bar(ax=ax, color='red', label='Closed')

title = 'Never operated reactors'
print(f'{title}:')
ax.set_title(title)
ax.set_xlabel(None)
format_years_ticks(ax)
ax.yaxis.set_major_formatter(lambda val, pos: int(val))
ax.text(0, ax.get_ylim()[0], copyright_text, fontsize=copyright_font_size, verticalalignment='bottom')
ax.legend()

fig.tight_layout()
plt.show()

Never operated reactors:

never_opened_df.groupby(['Country', 'Closed'])['Closed'].count().sort_values(ascending=False)

Country        Closed
Spain          1984.0    4
United States  1984.0    3
Belarus        1987.0    2
Cuba           1992.0    2
Germany        1990.0    2
Italy          1988.0    2
Poland         1990.0    2
Taiwan         2014.0    2
Ukraine        1987.0    2
               1990.0    2
United States  1983.0    2
Austria        1978.0    1
Germany        1985.0    1
Philippines    1986.0    1
Sweden         1970.0    1
Name: Closed, dtype: int64

never_opened_df.groupby(['Country', 'Begin building'])['Begin building'].count().sort_values(ascending=False)

Country        Begin building
United States  1975.0            4
Poland         1982.0            2
Ukraine        1988.0            2
               1984.0            2
Germany        1983.0            2
Italy          1982.0            2
Taiwan         1999.0            2
Belarus        1983.0            2
Spain          1972.0            2
               1975.0            2
Sweden         1965.0            1
Austria        1972.0            1
Philippines    1976.0            1
Germany        1972.0            1
Cuba           1985.0            1
               1983.0            1
United States  1977.0            1
Name: Begin building, dtype: int64

construction_time = input_df.groupby(['Begin building'])['Construction time'].agg(['min', 'median', 'max']).reindex(years, fill_value=0)

fig, ax = plt.subplots(figsize=(11, 5))

construction_time.plot.bar(ax=ax, width=0.8)

title = 'Years of construction time by the year when construction started - World'
print(f'{title}:')
ax.set_title(title)
ax.set_xlabel(None)
format_years_ticks(ax)
ax.yaxis.set_major_formatter(lambda val, pos: int(val))
ax.yaxis.set_major_locator(ticker.MultipleLocator(5))
ax.text(0, ax.get_ylim()[1] - 0.1, copyright_text, fontsize=copyright_font_size, verticalalignment='top')
ax.legend()

fig.tight_layout()
plt.show()

Years of construction time by the year when construction started - World:

top_countries = total_constructed_by_country['Country'].to_list()
construction_time = input_df.groupby(['Country', 'Begin building'])['Construction time'].agg(['min', 'median', 'max'])
ymax = construction_time.max(axis=None) + 1

for c in top_countries:
    df = construction_time[construction_time.index.get_level_values(0) == c].droplevel(0).reindex(years, fill_value=0)
    
    fig, ax = plt.subplots(figsize=(11, 5))

    df.plot.bar(ax=ax, width=0.8)

    title = f'Years of construction time by the year when construction started - {c}'
    print(title)
    ax.set_title(title)
    ax.set_xlabel(None)
    format_years_ticks(ax)

    ax.yaxis.set_major_formatter(lambda val, pos: int(abs(val)))
    ax.yaxis.set_major_locator(ticker.MultipleLocator(5))

    ax.set_ylim(0, ymax)
    ax.text(0, ax.get_ylim()[1], copyright_text, fontsize=copyright_font_size, verticalalignment='top')
    ax.legend()

    fig.tight_layout()
    plt.show()

Years of construction time by the year when construction started - United States

Years of construction time by the year when construction started - China

Years of construction time by the year when construction started - France

Years of construction time by the year when construction started - Japan

Years of construction time by the year when construction started - Russia

Years of construction time by the year when construction started - United Kingdom

Years of construction time by the year when construction started - Germany

Years of construction time by the year when construction started - India

Years of construction time by the year when construction started - South Korea

Years of construction time by the year when construction started - Canada

	Country	Plant name	Unit No.	Type	Model	Status	Capacity (MW)	Begin building	Commercial operation	Closed	Begin building date	Commercial operation date	Closed date	SubRegion	Region	Operated closed	Construction time
0	Argentina	Atucha	1	PHWR	Siemens-KWU	Operational	335.0	1968.0	1974.0	NaN	1968-06-01	1974-06-24	NaT	Latin America and the Caribbean	Americas	NaN	6.061602
1	Argentina	Atucha	2	PHWR	Siemens-KWU	Operational	692.0	1981.0	2014.0	NaN	1981-07-14	2014-06-27	NaT	Latin America and the Caribbean	Americas	NaN	32.952772
2	Argentina	Embalse	1	PHWR	CANDU-6	Operational	600.0	1974.0	1984.0	NaN	1974-04-01	1984-01-20	NaT	Latin America and the Caribbean	Americas	NaN	9.804244
3	Argentina	CAREM	1	PWR	CAREM25	Under construction	25.0	2014.0	NaN	NaN	2014-02-08	NaT	NaT	Latin America and the Caribbean	Americas	NaN	NaN
4	Armenia	Metsamor	1	PWR	VVER-440/V-270	Shut down	376.0	1969.0	1977.0	1989.0	1969-07-01	1977-10-06	1989-02-25	Western Asia	Asia	1989.0	8.265572
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
746	United States	Watts Bar	2	PWR	WH 4-Loop (ICECOND)	Operational	1164.0	1973.0	2016.0	NaN	1973-09-01	2016-06-04	NaT	Northern America	Americas	NaN	42.757016
747	United States	Wolf Creek	1	PWR	WH 4-Loop (DRY)	Operational	1200.0	1977.0	1985.0	NaN	1977-05-31	1985-09-03	NaT	Northern America	Americas	NaN	8.260096
748	United States	Yankee Rowe	1	PWR	WH (DRY)	Dismantled	167.0	1957.0	1961.0	1991.0	1957-11-01	1961-07-01	1991-10-01	Northern America	Americas	1991.0	3.663244
749	United States	Zion	1	PWR	WH (DRY)	Dismantled	1040.0	1968.0	1973.0	1998.0	1968-12-01	1973-12-31	1998-02-13	Northern America	Americas	1998.0	5.081451
750	United States	Zion	2	PWR	WH (DRY)	Dismantled	1040.0	1968.0	1974.0	1998.0	1968-12-01	1974-09-17	1998-02-13	Northern America	Americas	1998.0	5.793292

Table of contents¶

Number of constructed and operated reactors¶

World stats¶

Region stats¶

Subregion stats¶

Country stats¶

Never opened reactors¶

Construction time¶

	SubRegion	Region	Begin building	Commercial operation	Closed
0	Eastern Asia	Asia	181	147	34
1	Northern America	Americas	165	159	52
2	Western Europe	Europe	125	120	58
3	Eastern Europe	Europe	122	94	29
4	Northern Europe	Europe	67	64	45
5	Southern Asia	Asia	42	31	2
6	Southern Europe	Europe	21	15	13
7	Latin America and the Caribbean	Americas	11	7	2
8	Western Asia	Asia	10	5	1
9	Northern Africa	Africa	3	0	0
10	Sub-Saharan Africa	Africa	2	2	0
11	Central Asia	Asia	1	1	1
12	South-eastern Asia	Asia	1	0	1